Sonoelectrochemical Synthesis of Submicron Metal Powders
MASSACHUS F-rSN-WflTUTFE
OF TECHNOLOGY
Joseph Reneker
FEB 16201
B.S., Mechanical Engineering
B.S., Applied Physics
Purdue University, 2009
LIBRARIES
SUBMITTED TO THE DEPARTMENT OF MECHANICAL ENGINEERING IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
ARCHIVES
MASTER OF SCIENCE IN MECHANICAL ENGINEERING
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
FEBRUARY 2012
@2012 Joseph Reneker. All rights reserved.
The author hereby grants to MIT permission to reproduce
and to distribute publicly paper and electronic
copies of this thesis document in whole or in part
in any medium now known or hereafter created.
Signature of Author:
6/
Department of Mechanical Engineering
January 20, 2012
Certified by:
Taofang Zeng
Ao
Principal Research Scientist
,j
--j~
SiS Servisor
Accepted by:
David E. Hardt
Professor, Department of Mechanical Engineering
Graduate Officer, Department of Mechanical Engineering
11
2
Sonoelectrochemical Synthesis of Submicron Metal Powders
By
Joseph Reneker
Submitted to the Department of Mechanical Engineering
on January 19, 2012 in Partial Fulfillment of the
Requirements for the Degree of Master of Science in
Mechanical Engineering
ABSTRACT
Pulsed sonoelectrochemical synthesis is a widely used technique for producing
nanoparticles. In this technique, alternating pulses of electric current and power
ultrasound are applied to an electrochemical cell to create and suspend particles in the
electrolyte. The pulsed technique largely separates the particle morphology defining
physical action of electrochemistry and ultrasound. Despite the large body of work
characterizing the pulsed method, surprisingly little is written about the behavior of
particles in the continuous case, where electric current and ultrasound are
simultaneously present. In this thesis, continuous ultrasound assisted electrochemical
synthesis of nanoparticles is established. Potentially useful mechanisms for particle size
and shape control in continuous reactors are discussed.
A continuous
sonoelectrochemical reactor was designed and demonstrated to produce submicron
copper powders. Improvements to the batch reactor design are proposed to extend the
technique to a flow reactor useful for commercial production of submicron metal
powders.
Thesis Supervisor: Taofang Zeng
Title: Principal Research Scientist
4
Table of Contents
1.0 Introduction
2.0 Background
6
9
9
2.1 Electrochemistry
2.2 Electrodeposition of Powders
11
2.3 Ultrasonic Processing
11
2.4 Ultrasonic Electrodeposition
3.0 Experimental Apparatus Development
14
16
16
19
22
3.1 Initial Reactor Design
3.2 Improved Reactor Design
3.3 Improved Reactor Test
4.0 Discussion and Analysis
27
27
4.1 Continuous versus Pulsed
4.2 Particle Behavior in SonoelectrochemicaI
Reactors
28
4.3 Interactions on Electrode Surfaces
29
4.4 Interactions in Suspension
30
33
4.5 Suspensive Electrode
4.6 Continuous Flow Concept
5.0 Conclusions
5.1 Potential Applications
39
41
_________________________41
5.2 Future Development________________________
6.0 References
42
44
Table of Figures
1.1 - Schematic of operating sonoelectrochemical cell
7
2.1 - Schematic of acoustic streaming flow lines from an ultrasonic horn
13
3.1 - Initial sonoelectrochemical cell design
17
3.2 - Improved sonoelectrochemical cell design
18
3.3 - Histogram of particle diameters from improved reactor test
23
3.4 - TEM micrograph of submicron Cu particles depicting the smallest visible size
3.5 -TEM micrograph of submicron Cu particles depicting shape uniformity
25
4.1 - Schematic of the relative pulse timing for pulsed sonoelectrodeposition
28
4.2 - Schematic of cathode surface particle behaviors
31
4.3 - Schematic of dependence of particle adhesion with diameter
4.4 - Schematic of shape dependence of particle durability in ultrasonic fields
32
35
4.5 - Schematic of mechanically driven particle/particle interactions in suspension
36
4.6 - Schematic of suspensive electrode concept
37
4.7 -Schematic
37
of charge driven agglomeration
4.8 - Continuous flow sonoelectrochemical reactor concept
26
38
Acknowledgements
I would like to acknowledge the generous financial support provided to me by
Sandia National Laboratories. In addition, I would like to thank my advisor Dr. Taofang
Zeng for guidance in this project. Lastly, would like to thank my colleague Yi He for
answering my questions and providing constructive feedback on my ideas.
1.0 Introduction
Nanostructured materials can have unique size and shape dependent properties
which can differ greatly from their bulk form [1]. These properties are potentially useful in
medicine [2], catalysis [3], electronics [4], and countless other fields. The past several
decades have seen the development of a tremendous variety of techniques for the
production of nanomaterials, such as evaporation and condensation in low pressure inert
gases [5], combustion flame [6], laser ablation [7], chemical vapor deposition [8],
electrospray [9], and precipitation [10]. Despite their useful properties, and the myriad
production techniques, widespread commercial application of nanostructured materials
has been limited [11]. In order to speed adoption of such nanotechnology, even cheaper,
more flexible, and more reproducible methods of producing nanomaterials must be
developed.
This work concerns the sonoelectrochemical synthesis of submicron sized metal
powders. The sonoelectrochemical technique is based on a discovery [12] that
electroplating an operating ultrasonic horn, rather than evenly coating the horn's surface,
can instead yield a suspension of nanoparticles. Sonoelectrochemical methods promise to
be a technique for the low cost bulk production of nanostructured materials.
An example sonoelectrochemical cell is shown in Figure 1.1. At first glance, it
appears to be a typical copper plating electrochemical cell. Immersed in a bath of CuSO 4
are two electrodes, the anode and cathode. A power supply drives a current across the
cell. At the sacrificial anode, copper atoms are oxidized to Cu2+ ions. The ions migrate
through the electrolyte bath and are reduced on the cathode surface. Instead of a typical
stationary metal electrode, the cathode is a titanium ultrasonic horn. Before the copper
deposit can build up enough to coat the electrode, it is broken up and expelled into
suspension in the electrolyte. Understanding the mechanisms by which these deposits
grow, disperse, and behave in suspension is necessary to develop the technology for
eventual commercial application.
Sonoelectrochemistry is the intersection of several fields: electrochemistry, power
ultrasound,
and nanotechnology. Understanding how each of these disciplines
contributes to the process should enable the adjustment of experimental parameters to
tune the properties of the produced particles to suit. What follows is a general overview
of each technology, with particular attention paid to aspects important to understanding
sonoelectrochemical synthesis.
Power Supply
Ultrasonic
Horn/Cathode
VH
Aqueous
CuSO. Bath
Sacrificial
Anode
p
eq
0
*I
/
Figure 1.1- Schematic of operating sonoelectrochemical cell
2.0 Background
2.1 Electrochemistry
Electrodeposition of metals, also known as electrocrystallization or electroplating,
is a process where metal ions in solution are driven by an electric field to coat an
electrode surface. The morphology of the crystallized metal, known as the electrodeposit,
can vary dramatically on the particulars of the electrochemical setup. Electrochemical
cells can be designed to cover an electrode in a thin film of material, in a technique
known as electroplating. In general, interrelated parameters such as current density,
electrode potential difference, bath temperature, metal ion concentration, electrolyte
conductivity, bath additives, electrode material and geometry, and bath size all
contribute to the character of the deposit. A number of techniques have been developed
to produce nanostructured metal products by electrodeposition [13].
Copper was chosen as the product metal for this work because of the large
volume of literature concerning the production and characterization of nanostructured
copper materials. Typically, the electrodeposition of copper is governed by the following
reaction:
Cuz- + 2e- -> Cu
Dissolved copper ions are reduced on the cathode of an electrochemical cell to form
metallic copper. Though the reaction kinetics of many baths and additives are not fully
understood [14], copper electrodeposition is a well characterized industrial process. The
designation of experimental parameters is simplified by the large volume of available
trade and scientific literature.
The most widely used copper plating bath is based on an aqueous solution of
copper sulfate, CuSO 4 , with copper ion concentrations in the 10-60 g/L range [14]. In
solution, the solvated Cu 2 atoms are charge carriers, and eventually deposit on the
cathode by reduction. Higher metal ion concentrations enable faster mass deposition
[14], which can be useful for higher production rates. Addition of sulfuric acid, H2SO4 , to
the bath increases conductivity by introducing nonreactive charge carriers. Raising the
bath temperature tends to increase the conductivity of the electrolyte [15]. Increasing the
conductivity of a given bath tends to minimize potential gradients in the electrolyte,
which gives more uniform deposits [16].
The morphology of conventionally electrodeposited copper particles is well
understood to depend primarily on current density [17]. At all current densities, the
deposited particles have some degree of dendritic character. At relatively low current
densities, the dendrites are largest, broadest, and most branched. With increased current
density, the particles become flatter, with large crystalline surfaces. Even further
increased current densities yield different compact dendritic structures. Based on this
knowledge, copper appears a good choice for nanostructure production, as the large
deposits tend to have dendritic, and therefore generally more fragile structure.
2.2 Electrodeposition of Powders
Conventional electrodeposition has been used to produce metal powders
industrially for at least 80 years [18]. Commercially available powders produced by this
method have mean diameters in the 10-100 pm range [19]. There are several reported
techniques employed to control the size of these conventionally electrodeposited
powders. Generally, electrodeposited metals can form powders above the limiting
diffusion current [20]. Variations in the limiting current can be achieved by using a bath
with lower metal ion concentration, or by adding non reacting conductivity enhancers
[19]. Driving the electrochemical cells at over-potentials sufficient to decompose water
evolves hydrogen at the cathode, which can help loosen powder and yield dendritic type
powders [21]. Organic electrocrystallization inhibitors, such as chelating or complexing
agents [22] or surfactants [23], can be added to baths to reduce particle size. Removal of
electrodeposited powders can be done actively, by brushing [18] or by shocking the
powder off by rapping the cathode [21]. Alternatively, cathodes with low product metal
surface adhesion, such as graphite, passively shed powder under gravity [24].
2.3 Ultrasonic Processing
Ultrasound is generally defined as acoustical vibrations of more than 20 kHz. The
use of these waves can be divided into two main categories: diagnostic and power
ultrasound [25]. Diagnostic ultrasound generally encompasses applications where
ultrasonic waves do not dramatically change the sample. Medical ultrasonography
concerns the nondestructive transport of ultrasonic waves through tissue, and is an
example of diagnostic ultrasound. Power ultrasound, however, concerns applications
where ultrasound has intentional and often dramatic effects on the sample. Ultrasonic
welding couples acoustic vibrations to friction between two work pieces to join them.
Sonochemistry is another application of power ultrasound, where the application of
ultrasonic energy enables high temperature and pressure chemistry to be carried out with
great simplicity.
Several technologies have been developed to produce power ultrasound. The two
main types used in the laboratory setting are piezoelectric and magnetostriction.
Piezoelectric materials deform when electrically polarized. Stacks of piezoelectric material
are built and driven at ultrasonic frequencies to generate power ultrasound.
Magnetostriction is the tendency of a material to deform when magnetized. This is the
effect responsible for the 120 Hz hum of large transformers [26]. Ultrasonic generators
based on this principle magnetize and demagnetize material at an ultrasonic frequency.
There are several methods to couple the ultrasound to the experiment. Ultrasonic
cleaning baths accomplish this by fixing the generator to the floor or wall of the bath. In
an ultrasonic horn, the ultrasonic generator is fixed to a quarter or half wave resonator.
For strength and chemical stability these resonators are typically constructed from a
titanium alloy.
Power ultrasound provides two important effects to electrodeposition; cavitation
and acoustic streaming. An ultrasonic wave, like any other acoustic wave, is simply a
longitudinal wave that propagates by compression
and decompression.
When
propagating through a fluid, at high enough amplitude the pressure during the
decompression cycle can be lower than the vapor pressure of the fluid or gasses dissolved
in it. When this occurs, some of the dissolved gas or fluid vaporizes and forms bubbles, in
a process known as cavitation. As the pressure cycle continues, the gas bubble is
compressed. Eventually, sometimes after many compression cycles, the bubble will
collapse. This is a particularly violent process, as the temperature can routinely reach
5000 K [27]. This collapse generates shocks that can deform and erode nearby materials.
As a result, even ultrasonic resonators made of titanium tend to wear.
Acoustic streaming is the steady bulk flow of a liquid driven by the absorption of
acoustic waves. In power ultrasound applications, like ultrasonic cleaning baths, the flow
can be quite pronounced, amounting to a vigorous mixing action. Just like other
mechanical mixing, this bulk flow tends to minimize temperature and concentration
gradients [28]. A schematic representation of the acoustic streaming flow lines is included
in Figure 2.1.
Bulk fluid
flow
Figure 2.1- Schematic of acoustic streaming flow lines from an ultrasonic horn
2.4 Ultrasonic Electrode position
The application of ultrasound to electrodeposition dates back to at least 1936,
with the publication pioneering work on the subject [29]. These researchers found that
when a plating bath was subject to ultrasonic waves, a rippling pattern with the same
wavelength was formed in the electrodeposit. This basic work proved that ultrasound
changes reaction kinetics of electrodeposition. Subsequent studies of the interaction
found that the application of ultrasound can improve the industrial electrodeposition
process. Several reviews of the topic have been published [30] [31]. Though most of the
work concerns topics of utility to industrial users, such as electrode geometry and current
densities, there are many results useful to users interested in the production of
nanostructured materials.
The effects of cavitation from the application of ultrasound are the most
apparent. As described earlier, the formation and implosion of gas bubbles can violent
localized shocks, which can dislodge nearby adhered material and gas bubbles. When
electrodeposition conditions favor dendrite growth, cavitation is known to break off
perpendicular growths and inhibit dendrite formation, instead favoring dense crystal
growth [25]. When a cavity moves or collapses near a surface, it displaces fluid and
creates rapidly moving streams near the surface [32]. This is important from the
perspective of electrodeposition, because the effect causes bulk transport of fluid from
outside the conventional double layer. Effectively, this increases the maximum rate at
which ions can be transported to the surface for reduction because the process is no
longer diffusion limited [30]. For purposes of calculation, this can be thought of as a
decrease in the size of the Nernst diffusion layer.
3.0 Experimental Apparatus Development
3.1 Initial Reactor Design
Initial experiments were run on an electrochemical cell provided by a colleague.
The cell was operated in a 500 mL glass beaker. A cylindrical sheet of 18 gauge bare
copper was used as an anode. The tip of a Y2" diameter solid tip Ti-6Al-4V titanium
ultrasonic horn was used as a cathode. Power was supplied to the cell by an Agilent
U8002A power supply, which could be operated in both galvanostatic and potentiostatic
modes. Stranded copper wire soldered to the anode provided positive electrical
connection to the power supply. The titanium cathode was connected to the power
supply by twisting a stranded copper wire around the base of the Y2" diameter section of
the horn. 20 kHz Ultrasonic energy was supplied to the horn by a VCX500 500 watt
piezoelectric PZT ultrasonic transducer. The electrochemical cell was temperature
controlled by immersing the base in a bath supplied by the external output of a
conventional thermostated laboratory water bath. An image of this cell is provided in
Figure 3.1. This image was taken after an experimental run, and depicts the cathode after
it was removed from the electrolyte.
Preparing and running the electrodeposition experiment was relatively simple.
First, the electrolyte bath was prepared by dissolving a quantity of copper sulfate in the
required volume of water, typically 250 mL. After transferring this transparent blue
electrolyte to the bath, the thermostat was activated and the cell brought to thermal
equilibrium. The ultrasonic horn was positioned such that the tip was approximately 1 cm
Figure 3.1 - Initial sonoelectrochemical cell design
Figure 3.2 - Improved sonoelectrochemical cell design
20
immersed in the electrolyte, giving an immersed surface area of 1.3 cm2 . Then, the horn
was activated and the output power was adjusted high enough such that cavitation just
started to appear. Then, the power supply was activated. After around 20 minutes, the
electrolyte would gradually start to turn an opaque green. This color indicated that large
quantities of copper particles were in suspension. At this point, the power supply was
turned off, then the ultrasonic power.
Samples of the suspended powder were collected using a centrifugation
technique. First, the electrolyte solution was decanted from the electrochemical cell to
remove the largest copper particles. Then, the solution was poured into 50 mL plastic
centrifuge tubes and centrifuged for 10 minutes. After removal from the centrifuge, the
now clear blue electrolyte solution was carefully poured out, leaving a thin but visible
coating of copper particles on the surface of the centrifuge tube. The tubes were then
filled with a convenient fluid, such as water or ethanol, and shaken to return the particles
into suspension. TEM samples were prepared by dropping a drop of suspended particles
on a carbon coated copper mesh TEM grid from SPI. These samples were then
characterized by a Delong LEVM5 desktop SEM/TEM.
3.2 Improved Reactor Design
The initial reactor had many design flaws. Firstly, the ultrasonic horn, being shared
equipment, had a somewhat eroded tip. Tip erosion is an unavoidable process when an
ultrasonic horn is used to produce cavitation bubbles. Instead of being an ideal flat
surface, the tip of the horn was pitted and rough. This reduces the efficiency of the
ultrasonic horn, as well as promoting fouling, as copper would entrain in the pitted
surface and become difficult to remove. The cylindrical portion of the horn tended to
entrain copper as well. Figure 3.1, which depicts the initial cell immediately after use,
shows the copper fouling of the cylindrical portion of the titanium horn. Additionally, the
continuity of the electrical contact for the cathode was suspect, as the hookup wire
wasn't securely fixed to the cathode.
Several improvements to the cathode were made. The updated design is pictured
in Figure 3.2. First, as the titanium horn was damaged irreparably, a new resonator was
purchased. The new horn had a removable and replaceable tip, which means it will be
possible to refurbish it in the future when the tip becomes worn. In order to reduce the
theoretical problem of deducing the mechanisms of cathode particle formation and
separation, as well as to bring the work in line with most of the reported
sonoelectrochemical literature, the problem of insulating the cylindrical portion of the
cathode was addressed. Though most papers mention that the cylindrical wall of the
cathode was insulated [33], the literature is often ambiguous about the precise
construction of this insulation. After several attempts it was found that a section of heat
shrink tubing, shrunk around the immersed part of the cathode, worked well. It's worth
pointing out that placing solid materials in contact with the titanium horn tends to damp
the resonator, so it's best to minimize the amount of heat shrink tubing used. The last
improvement to the cathode was the electrical connection to the power supply. While it
would have been better to have a solid connection, it was discovered that soldering to
titanium is nontrivial. The new ultrasonic horn was purchased with an adaptor for a
removable reaction vessel. When the adaptor was installed, it was possible to assemble
the instrument with the cathode wire squeezed in place. The connection was robust
enough that it didn't require tightening after every run, like the twisted wire did on the
initial cathode.
Improvements were made to the anode as well. In the initial design, the anode
was constructed from a piece of copper sheet. Chosen for the sake of symmetry, the
cylindrical anode was positioned in concentric with the cathode. The anode sat tight
against the walls of the glass reaction vessel. The immersed surface area of this electrode
was approximately 80 cm2 , about 60 times larger than the anode area. It was found that
during operation, the electrode tended to trap a thin layer of electrolyte between it and
the surface of the glass. Layers like decrease the thermal conductivity of the wall, possibly
raising the temperature of the electrolyte. As such, a much smaller anode was designed
for the improved reactor. This anode, also made of copper, was circular and for the sake
of symmetry was placed underneath the cathode. Electrical connection to the power
supply was made by soldering a stranded copper wire to the back of the plate. This wire
allowed the anode to be easily repositioned. Another consideration for the anode was
whether to make it sacrificial or passive. Sacrificial anodes can be useful for maintaining
metal ion concentration in the electrolyte, ideally where one atom is ionized for every ion
that is passivated. However, attention must be paid to the relative efficiencies of the
anode and cathode. At sufficiently large over-potentials, side reactions such as the
decomposition of water can occur. Differences in electrode reaction efficiencies can
cause metal ion concentrations to increase [16]. Such imbalances may disturb carefully
controlled industrial or laboratory electrodeposition reactions.
The last improvements were made to the cell itself. The initial reaction vessel was
an open beaker, which allowed electrolyte to spray out and evaporate freely.
Additionally, the reaction vessel wasn't mechanically coupled to the ultrasonic horn
assembly, which meant that the relative electrode geometry wasn't set. During operation,
the glass beaker would occasionally drift, driven by currents in the cooling bath or the
coupled oscillations from the horn. Additionally, the initial reactor vessel was large, given
the minute quantity of samples required. Therefore, the new vessel was chosen such that
a smaller volume of electrolyte could be easily processed. The new reaction vessel was
supplied by the company that produced the ultrasonic horn, Sonics and Materials, Inc.
With a supplied adaptor, this vessel attached to the ultrasonic horn, enabling more
precise control of the relative electrode positions. The vertical position of the horn was
adjustable in the reaction vessel, so the volume of electrolyte to be processed could be
adjusted to suit.
3.3 Improved Reactor Test
The following is an example run, which demonstrates the ability of the machine to
produce nanoparticles. 250 mL of 0.5 M CuSO 4 was prepared and loaded into the
sonoelectrochemical cell. The cell was placed in a water bath and kept temperature
stable at 30 C. The ultrasonic horn was activated, and the output was adjusted until
cavitation just started to occur. Then, the power supply was activated in potentiostatic
mode at 1.1 V. Over the course of 20 minutes, the current rose from an initial value of
0.21 A to 0.43 A. After 20 minutes, the power supply and the ultrasonic horn were turned
off, in that order. The particle suspension was centrifuged and samples prepared for the
TEM. Figure 3.4 and Figure 3.5 are TEM images of the particles. The images show a range
of particle sizes, with many particles in the sub 100 nm range. Images were processed in
ImageJ, and a histogram of particle diameters was prepared, presented here as Figure
3.3. The size distribution of the particles is broad, but weighted heavily towards below
100 nm.
25
-
Particle Size Distribution
20 -
15
-
10
-
iI.i.U
Diameter (nm)
Figure 3.3 - Histogram of particle diameters from improved reactor test
S1
um
I
Figure 3.4 - TEM micrograph of submicron Cu particles depicting smallest visible size
27
2 um
Figure 3.5 - TEM micrograph of submicron Cu particles depicting shape uniformity
28
4.0 Discussion and Analysis
4.1 Continuous versus Pulsed
A review of the sonoelectrochemical literature finds practically all work concerns
pulsed sonoelectrodeposition; that is, where the ultrasonic and electroplating processes
are both pulsed and out of phase. A schematic representation of the pulsed process is
included in Figure 4.1. The typical pulsed sonoelectrodeposition process goes as follows:
A cycle starts with a clean electrode. During the electroplating step, an electric current
pulse is activated and the surface of the electrode becomes covered in particle nuclei. As
the electroplating pulse progresses, the nuclei grow larger. At the end of the electric
pulse, the nuclei have grown to their final size. After the electric current is turned off, an
ultrasound pulse removes the particles from the electrode and expels them into
suspension in the electrolyte. The cycle starts over with the newly cleaned electrode.
The pulsed method has been used to produce highly monodisperse nanoparticles
from many metals and semiconductors, including Cu [34], Pt [35], Mg [36], and CdSe [37].
This technique is pleasing from an analytical standpoint. By neatly separating the
electrochemical and ultrasonic processes, one can leverage large bodies of previous work
which have characterized much of the underlying physics of both processes. The
distribution of particle sizes reported in this work is inferior to samples prepared using
the pulsed technique [38]. One could argue that this result indicates the continuous
method is inferior to the pulsed method. However, the pulsed method precludes
capitalizing on unique physics which occur when electrochemistry and ultrasound are
present simultaneously. Improvements to the continuous design might be made with a
better physical understanding of the interactions present in both types of reactors. What
follows is a discussion of effects which occur in both pulsed and continuous
sonoelectrochemical reactors, as well as some that are only present in the continuous
case.
Particle Separation
Ultrasonic
Power
Electric
Current
Nucleation and Growth
Time
Figure 4.1- Schematic of the relative pulse timing for pulsed sonoelectrodeposition
4.2 Particle Behavior in Sonoelectrochemical Reactors
In sonoelectrochemical reactors, particle shapes and sizes can be modified by
electrochemical or mechanical forces. What follows is a qualitative discussion of some of
the mechanisms by which particles in a sonoelectrochemical cell could interact with one
another and with the components of the device itself. Methods for minimizing the effects
are discussed, with the goal of improving future continuous sonoelectrochemical
reactors.
4.3 Interactions on Electrode Surfaces
The behavior of particles on the surface of the cathode is of great interest.
Particles form on the surface of the cathode by spontaneous nucleation. In this process,
metal ions are reduced on the surface of the cathode. The atoms travel across the surface
and coalesce into nuclei, which then grow into larger particles. The particles themselves
can coalesce into larger particles on the surface. Schematics of these concepts are
depicted in Figure 4.2. As these processes are continuous, in general there is a surface
inventory of particles of various sizes. The times particles spend on the cathode
determine their size; namely, particles which reside longer grow larger. Application of
ultrasound alters the kinetics of these processes [30]. The rate at which these processes
occur depends on the electrochemical properties of a given reactor, though they are
commonly controlled with additives such as chelating agents or surfactants [14].
The application of ultrasound expels particles from electrodes into suspension in
the electrolyte. Identifying general trends in this behavior will be useful for the design of
future continuous sonoelectrochemical reactors. One such trend is schematically
illustrated in Figure 4.3. There are several possible mechanisms for particle separation,
though large local shock from ultrasonic cavity collapse is often regarded as the primary
driver [31]. The ability of a particle to adhere to the cathode is likely size dependent. In
the minimum dimensional limit, single adsorbed atoms are not dislodged by cavitation
shocks. In the large particle limit the particles are subject to bulk fluid motion and inertial
effects and are more susceptible to removal by ultrasound. Therefore, there should be
some critical size where particles eventually dislodge.
4.4 Interactions in Suspension
Ultrasonic energy is also coupled to particles which come into the vicinity of the
ultrasonic horn tip where cavitation occurs. Just as at the horn surface, cavitation
produces highly localized and powerful shocks, which can break particles up. This concept
is illustrated in Figure 4.4. The magnitude of this effect depends on the particle shape.
Ultrasound tends to break up long dendritic structures, such as whiskers, more effectively
than compact crystalline structures, like round particles [25].
Particles can also interact with one another in the bulk electrolyte. Two such
situations are depicted in Figure 4.5. In the first case, particles collide in suspension and
agglomerate or coalesce into a single larger particle. This process is known as Ostwald
Ripening [39], and is often observed as an aging mechanism in stationary suspensions of
nanoparticles. The rate at which particles combine by this process decreases with particle
radius. The second particle interaction mechanism is collisions which yield additional
smaller particles. Here, particles strike one another with sufficient center of mass energy
to disintegrate into additional particles. While this process depends on the particle shape,
structure, composition, and various fluid properties, it likely dominates Ostwald ripening
for larger particles. This mechanism could be the mechanism for mean particle size
reduction observed in an experiment [40] where a nanoparticle suspension was sonicated
for 24 hours.
Nucleation and growth
Surface coalescence
Figure 4.2 - Schematic of cathode surface particle behaviors
Before Ultrasound
After Ultrasound
Small Particles
0
Large Particles
Figure 4.3 - Schematic of dependence of particle adhesion with diameter
4.5 Suspensive Electrode
The suspensive electrode concept [40] is depicted in Figure 4.6. Here, uncharged
particles from suspension come into contact with an electrode and pick up a net charge.
After returning to solution, the particles behave as the electrode they came into contact
with as they return to an uncharged state. For a bath with positive metal ions, particles
which come into contact with the cathode acquire net negative charge, and reduce
positive metal ions the electrolyte and grow in size. Particles which come into contact
with the anode decrease in size as they shed charge by releasing oxidized ions into the
electrolyte. In general, the rates of these processes may be asymmetric, depending on
the electrochemistry of the particular bath. For example, in a system where the rate
reduction of dissolved ions onto a charged particle is greater than the rate of oxidation of
copper atoms from a positively charged particle, the system would tend to increase the
size of particles over time. Experiments have shown that these highly local
electrochemical reactions proceed at a faster rate than other particle growth mechanisms
[40].
A corollary to the suspensive electrode theory concerns other behaviors of the
suspended charged particles. As the particles carry charge, they carry electric current.
Suspended particles have been shown to contribute significantly to the conductivity of an
electrochemical cell when ionic charge carriers have been depleted [40]. There is also a
particle-particle interaction mechanism. Particles which become charged on one
electrode can come into contact with a particle which was charged on the other
electrode. The two particles combine and the new, larger particle has a smaller net
charge. A schematic of this process is depicted in Figure 4.7. The rate at which this
process occurs relative to others can be understood as follows. The characteristic time
which a particle remains charged can be related to the concentration of active ions in the
electrolyte, as this is the mechanism for particle charge shedding in the bulk electrolyte.
High ion concentration means a particle will react and neutralize faster than a particle in
an electrolyte with low ion concentration. The characteristic charge neutralization time
can be related to some characteristic particle velocity in the sonoelectrochemical cell.
Recall that acoustic streaming tends to increase the average velocity of fluid in the
system. If the electrodes are far away, oppositely charged particles will tend to neutralize
before coming into contact in suspension.
While these processes appear symmetric, it is important to understand the
asymmetries introduced to conventionally designed sonoelectrochemical cells; namely
the dual use of the cathode as ultrasonic oscillator. A consequence of this design is the
relative intensity of ultrasound on the anode and cathode. Cavitation, and the
concomitant dramatic mechanical effects, is typically constrained to the tip of the
ultrasonic horn and cathode. The second asymmetry occurs by choice of deposited
material. Electrodeposition of cations, like Cu2+, occurs on the cathode. Nucleated
particles and cavitation only occur on the cathode in this system.
Before Ultrasound
After Ultrasound
Whiskers
Round Particles
Fcoy
Figure 4.4 - Schematic of shape dependence of particle durability in ultrasonic fields
Kinetic agglomeration or free
coalescence
Kinetic particle breakup
Figure 4.5 - Schematic of mechanically driven particle/particle interactions in suspension
Cathode
e
Anode
I
0
Figure 4.6 - Schematic of suspensive electrode concept
Figure 4.7 - Schematic of charge driven agglomeration
Reactor
Volume
Electrolyte
Particle
Suspension
Figure 4.8 - Continuous flow sonoelectrochemical reactor concept
4.6 Continuous Flow Concept
Based on these predictions, continuous sonoelectrochemical reactors should
function in the following manner. In operation, the reactor nucleates a large quantity of
small particles on the cathode. Once the particles coalesce and grow to a certain critical
size, they tend to be ejected into suspension in the electrolyte. The suspensive electrode
effect tends to enlarge the particle over time. Ostwald ripening tends to increase the size
of the particles, and is a function of time. The longer a batch of electrolyte is run, the
larger the first produced particles get relative to later produce particles. This leads to a
general broadening the size distribution. Therefore, in order to produce a more uniform
nanoparticle product, the effect of these interactions should be minimized.
A continuous reactor concept is proposed to improve the quality of the
nanoparticles created by continuous sonoelectrodeposition. A schematic of this concept
is depicted in Figure 4.8. To minimize many of the electrochemical and mechanical
particle interactions, the particles should be removed from the reactor as quickly as
possible after creation. This can be accomplished by using a small sonoelectrochemical
reaction volume, with an inlet for fresh electrolyte, and an outlet for product stream.
Given a steady flow to and from the reactor, an average residence time can be defined for
a fluid volume or a given particle in the reactor. A small residence time minimizes many of
the detrimental particle size distribution broadening effects. After removal from the
reactor, other stabilization techniques, which may or may not be compatible with the
conditions in the reactor, can be used, such as addition of antioxidants or centrifugation
and drying.
5.0 Conclusions
5.1 Potential Applications
The active components in for sonoelectrochemistry are relatively cheap, often
priced below $10,000 for ultrasonic generators which can produce more than 500 watts
of radiated ultrasonic power. As such, powders produced by this method would likely be
less expensive than those produced by other methods, particularly those that rely on high
purity reagents, such as sol-gel or colloidal synthesis, or complex equipment, like
chemical vapor deposition or lithographic techniques. While this technique probably
won't ever develop highly monodisperse distributions of particles due to the continuous
nature of particle formation and growth, there are still potential commercial applications.
This technique is potentially useful for the bulk production of submicron metal powders
where a wide particle size distribution can be tolerated.
An example of such an application could be a new technique for the low volume
production of printed circuit boards (PCBs). Many conventional PCB manufacturing
methods rely on masking and chemical etching of copper clad dielectric sheets [14]. Such
techniques utilize chemical baths, which are often dangerous and require specialized
industrial facilities. As such, low volume multilayer PCBs required for prototyping or
laboratory use are often expensive and with long lead times.
An alternative technique which required less complicated equipment would be
valuable. One proposed method would be to print silver particle impregnated fluids
directly onto dielectric boards using conventional ink-jet printing techniques [4]. High
temperature treatment of these boards would evaporate or react away the carrier fluids,
and sinter the silver particles into conductive tracings. In this case, the size distribution of
the silver powders is less important than low material cost.
5.2 Future Developments
More advanced electrodeposition techniques, such as two layer baths, haven't
been explored as a technique for reducing the reactivity of particles in suspension and as
a technique for removing them from the primary reactor bath. Two layer baths often
consist of an aqueous layer, where the electrodeposition occurs, and a nonpolar organic
or surfactant layer, where particles tend to accumulate [41]. Two layer baths can isolate
the electrically active polar layer from the atmosphere, which can lead to higher
electrode efficiencies [42].
Commercialization of this technique will require solving several outstanding
problems. First there isthe problem of removing the product particles from suspension in
the electrolyte. This work has relied on centrifugation and washing of small volumes of
particle suspensions, which is an inherently batch technique. Alternative methods would
need to be developed in order to make this process truly continuous.
Additionally, the long term suitability of these techniques has not been studied.
Most studies reported in the literature concern reactors that operate on the scale of
hours. In order to scale up these procedures, the long term behavior of components
needs to be studied in the context of continuous reactor operations. Problems might
arise, for example, from fouling of the electrodes with material created in side reactions.
Another problem might be erosion of the tip of the ultrasonic generator from cavitation.
This would lead to minor contamination of the product with titanium, as well as
continuous monitoring of the reactor components for wear. While these are potential
problems, wear and fouling are common issues with the scale up of laboratory processes.
A complete understanding of the underlying physics of sonoelectrodeposition is
still elusive. The combined physical effects of continuous electrodeposition and power
ultrasound may still yield new and useful nanostructured products. Further work should
be done to interrogate these effects. Advances in understanding of the continuous case
will have applications for pulsed sonoelectrodeposition, and vice versa.
6.0 References
1. Puntes, V. F., Krishnan, K. M., & Alivisatos, A. P. (2001). Colloidal nanocrystal shape
and size control: the case of cobalt. Science, 291, 2115-2117.
2. Rosi, N. L., & Mirkin, C.A. (2005). Nanostructures in Biodiagnostics. Chemical Reviews,
105, 1547-1562.
3. Bell, A. T. (2003). The impact of nanoscience on heterogeneous catalysis. Science, 299,
1688-1691.
4. Osch, T. H. J., Perelaer, J., Laat, A. W. M., & Schubert, U. S. (2008). Inkjet printing of
narrow conductive tracks on untreated polymeric substrates. Advanced Materials, 20,
343-345.
5. Ward, M. B., Brydson, R., & Cochrane, R.F.(2006). Mn nanoparticles produced by
inert gas condensation. Journal of Physics: Conference Series, 26, 296-299.
6. Violi, A., & Venkatnathan, A. (2006). Combustion-generated nanoparticles produced in
a benzene flame: a multiscale approach. The Journal of Chemical Physics, 125,
054302.
7. Makimura, T., Mizuta, T., Murakami, K. (2000). Formation dynamics of silicon
nanoparticles after laser ablation studied using plasma emission caused by secondlaser decomposition. Applied Physics Letters, 76, 1401-1403.
8. Swihart, M. T. (2003). Vapor-phase synthesis of nanoparticles. Current Opinion in
Colloid and Interface Science, 8, 127-133.
9. Nakaso, K., Han, B., Ahn, K. H., Choi, M., & Okuyama, K. (2003). Synthesis of nonagglomerated nanoparticles by an electrospray assisted chemical vapor deposition
(ES-CVD) method. Aerosol Science, 34, 869-881.
10. Shi, J., & Verweij, H. (2005). Synthesis and purification of oxide nanoparticle
dispersions by modified emulsion precipitation. Langmuir, 21, 5570-5575.
11. Malsch, I. (1997). Nanotechnology in europe: experts' perceptions and scientific
relations between sub-areas. Sevilla: Institute for Prospective Technological Studies.
12. Reisse, J., Caulier, T., Deckerkheer, C., Fabre, 0., Vandercammen, J., Delplancke, J. L.,
et al. (1996). Quantitative sonochemistry. Ultrasonics Sonochemistry, 3(3), S147-S151.
13. Bicelli, L.P., Bozzini, B., Mele, C., & D'Urzo, L.(2008). A review of nanostructural
aspects of metal electrodeposition. International Journal of Electrochemical Science,
3, 356-408.
14. Reid, J.(2000). Copper electrodeposition: principles and recent progress. Japanese
Journal of Applied Physics, 40, 2650-2657.
15. Kuyucak, S., & Chung, S. H. (1994). Temperature dependence of conductivity in
electrolyte solutions and ionic channels of biological membranes. Biophysical
Chemistry, 52, 15-24.
16. Takahashi, K. M. (2000). Electroplating copper onto resistive barrier films. Journal of
The Electrochemical Society, 147(4), 1414-1417.
17. Maksimovic, V., Pavlovic, Lj., Pavlovic, M., & Tomic, M. (2009). Characterization of
copper powder particles obtained by electrodeposition. The Journal of Metallurgy,
15(1), 19-27.
18. Drumiler, D. W., Moulton, R.W., & Putnam, G. L. (1950). Electrodeposition of copper
powder from acid sulfate baths. Industrial and Engineering Chemistry, 42(10), 20992102.
19. Delplancke, J. L., Dille, J., Reisse, J., Long, G.J., Mohan, A., & Grandjean, F. (2000).
Magnetic nanopowders: ultrasound-assisted electrochemical preparation and
properties. Chemistry of Materials, 12, 946-955.
20. Popov, K. I., Pavlovic, M. G., & Maksimovic, M. D. (1982). Comparison of the critical
conditions for initiation of dendritic growth and powder formation in potentiostatic
and galvanostatic copper electrodeposition. Journal of Applied Electrochemistry, 12,
525-531.
21. Osborn, W. H., & Tuwiner, S.B. (1944). Electrodeposition of copper powder.
Transactions of the Electrochemical Society, 85, 107-118.
22. Chu, C. M., & Wan, C.C.(1992). The effect of chelating agents on the cathodic
polarization and the electrodeposition of iron powders. Journal of Materials Science,
27, 6700-6706.
23. Gomes, A., & Pereira, M. 1.S. (2006). Pulsed electrodeposition of Zn in the presence of
surfactants. Electrochimica Acta, 51, 1342-1350.
24. Fil, T. I., & Khimichenko, Yu. 1.(1988). Production of ultrafine metal powders by
electrodeposition on a graphite cathode. Poroshkovaya Metallurgiya, 7, 6-9.
25. Mason, T. J., & Lorimer, J. P. (2001). Applied sonochemistry: the uses of power
ultrasound in chemistry and processing. Weinheim: Wiley-VCH.
26. Weiser, B., Pfuetzner, H., & Anger, J. (2000). Relevance of magnetostriction and forces
for the generation of audible noise of transformer cores. IEEE Transactions on
Magnetics, 36(5), 3759-3777.
27. Flint, E. B., & Suslick, K.S. (1991). The temperature of cavitation. Science, 20, 13971399.
28. Richardson, P. D. (1967). Heat transfer from a circular cylinder by acoustic streaming.
Journal of Fluid Mechanics, 30(2), 337-355.
29. Young, W. T., & Kersten, H. (1936). An effect of ultrasonic radiation on
electrodeposits. The Journal of Chemical Physics, 4, 426-427.
30. Kochergin, S. M., & Vyaseleva G.Y. (1966). Electrodeposition of metals in ultrasonic
fields. New York: Consultants Bureau.
31. Kapustin, A. P., Trofimov, A. N., & Belayev, L. M. (1970). Electrocrystallization of
metals in an ultrasonic field. Jerusalem: Israel Program for Scientific Translations.
32. Elder, S. A. (1958). Cavitation microstreaming. The Journal of the Acoustical Society of
America, 31(1), 54-64.
33. Reisse, J., Francois, H., Vandercammen, J., Fabre, 0., Kirsch-de Mesmaeker, A.,
Maerschalk, C., et al. (1994). Sonoelectrochemistry in aqueous electrolyte: A new type
of sonoelectroreactor. Electrochimica Acta, 39(1), 37-39.
34. Haas, I., Shanmugam, S., & Gedanken, A. (2006). Pulsed sonoelectrochemical
synthesis of size-controlled copper nanoparticles stabilized by poly(Nvinylpyrrolidone). The Journal of Physical Chemistry B, 110(34), 16947-16952.
35. Shen, Q., Jiang, L., Zhang, H., Min, Q., Hou, W., & Zhu, J. (2008). Three-dimensional
dendritic pt nanostructures: Sonoelectrochemical synthesis and electrochemical
applications. The Journal of Physical Chemistry C, 112(42), 16385-16392.
36. Haas, I., Shanmugam, S., & Gedanken, A. (2008). Synthesis of copper dendrite
nanostructures by a sonoelectrochemical method. Chemistry - A European Journal,
14(15), 4696-4703.
37. Shen, Q. M., Jiang, L., Miao, J., Hou, W., & Zhu, J.J.(2008). Sonoelectrochemical
synthesis of CdSe nanotubes. Chemical Communications, 1683-1685.
38. Saez, V., & Mason, T. J.(2009). Sonoelectrochemical synthesis of nanoparticles.
Molecules (Basel, Switzerland), 14(10), 4284-4299.
39. Voorhees, P. W. (1985). The theory of ostwald ripening. Journal of Statistical Physics,
30(1), 231-252.
40. Socol, Y., Abramson, 0., Gedanken, A., Meshorer, Y., Berenstein, L., & Zaban, A.
(2002). Suspensive electrode formation in pulsed sonoelectrochemical synthesis of
silver nanoparticles. Langmuir, 18(12), 4736-4740.
41. Myalkovskii, V. V., Shvets, T. M., Vasilenko, V. P., & Mel'nichenko, Z. M. (1984).
Structure of particles of very finely divided electrolytic iron. Poroschkovaya
Metallurgiya, 5(257), 1-6.
42. Balinka, M. N., Serpuchenko, E.A., & Kurilenko, 0. D. (1983). Production of copper
powders in a two-layer electrolytic bath in the presence of higher fatty acids.
Poroshkovaya Metallurgiya, 7(259), 1-4.